CDDRL WORKING PAPERS. The Quality of Democracy. Larry Diamond Leonardo Morlino. Number September 2004

Transcription

1 CDDRL WORKING PAPERS The Quality of Democracy Larry Diamond Leonardo Morlino Center on Democracy, Development, and The Rule of Law Stanford Institute on International Studies Number September 2004 This working paper was produced as part of CDDRL s ongoing programming on economic and political development in transitional states. Additional working papers appear on CDDRL s website:

2 Center on Democracy, Development, and The Rule of Law Stanford Institute of International Studies Stanford University Encina Hall Stanford, CA Phone: Fax: About the Center on Democracy, Development and the Rule of Law (CDDRL) CDDRL was founded by a generous grant from the Bill and Flora Hewlett Foundation in October in 2002 as part of the Stanford Institute for International Studies at Stanford University. The Center supports analytic studies, policy relevant research, training and outreach activities to assist developing countries in the design and implementation of policies to foster growth, democracy, and the rule of law. About the Author Larry Diamond is a senior fellow at the Hoover Institution. He is also professor of political science and sociology (by courtesy) at Stanford University and coordinator of the Democracy Program of the new Center for Democracy, Development, and the Rule of Law at Stanford's Institute for International Studies. Diamond is a specialist on democratic development and regime change and on U.S. foreign policy affecting democracy abroad. His current scholarly research examines comparative trends in the quality and stability of democracy in developing countries and postcommunist states, public opinion in new democracies (especially in East Asia), and democratic development in Taiwan. He also writes and consults on U.S. international efforts to promote democracy abroad and on international development policies and programs. Leonardo Morlino is the Professor of Political Science at the University of Florence (Italy) and director of the Research Centre on Southern Europe. He is a visiting professor at Stanford University (USA), Nuffield College (Oxford, UK), Institute d études politiques (Paris, France), Hoover Institution (USA), Center for Advanced Research in Social Sciences (Juan March, Madrid, Spain), and Yale University (USA). His publications include more than hundred articles in Italian, French, English, Spanish, and German mainly on authoritarianism and democratic theory. He is also the author of several books such as Come cambian los regimenes politicos (Madrid 1985), Democracy Between Consolidation and Crisis. Parties, Groups, and Citizens in Southern Europe (Oxford, 1998), Democrazie e Deocratizzazioni (Bologna 2003).

3 The Quality of Democracy By Larry Diamond and Leonardo Morlino As democracy has spread over the past three decades to a majority of the world s states, analytic attention has turned increasingly from explaining regime transitions to evaluating and explaining the character of democratic regimes. Much of the democracy literature of the 1990s was concerned with the consolidation of democratic regimes. 1 In recent years, social scientists as well as democracy practitioners and aid agencies have sought to develop means of framing and assessing the quality of democracy. This stream of theory, methodological innovation, and empirical research has three broad motives: first, that deepening democracy is a moral good, if not an imperative; second, that reforms to improve democratic quality are essential if democracy is to achieve the broad and durable legitimacy that marks consolidation; 2 and third, that long-established democracies must also reform if they are to attend to their own gathering problems of public dissatisfaction and even disillusionment. In fact, these latter trends the broad decline of public confidence in governmental and political institutions, the growing citizen alienation from political parties in particular, and the widespread perceptions that democratic governments and politicians are increasingly corrupt, self-interested and unresponsive are common to many democracies, new and old, and have even led prominent researchers to speak of a crisis of democracy. 3 Touching as it does on basic questions of norms and legitimacy, the quality of democracy is a value-laden and hence controversial subject. Who is to define what constitutes a good democracy, and to what extent is a universal conception of democratic quality possible? How can the effort to address avoid becoming paternalisticexercises in which the established democraciestake themselves for granted as models and so escape scrutiny? How can assessments of 1

4 democratic quality go beyond mere analytics and be useful to political reformers, civil society activists, international donors, and others who seek to improve the quality of democracy? These are only some of the questions that pervade and motivate this growing sub-field of study. The five thematic essays that follow are part of a collaborative effort, launched at a conference at Stanford University, to elaborate and refine the concept of democratic quality and to apply it to a series of six paired comparisons. The concluding essay in the present collection, by Marc Plattner, develops critical comments he offered at that conference. 4 The authors of the five thematic essays were each asked to define and articulate a particular dimension of the quality of democracy, to explain how it relates to other dimensions in our framework, to suggest possible indicators for measuring the dimension, to identify ways in which this element of democratic quality is subverted in the real world (and how these subversions can be empirically detected), and to offer (where possible) policy recommendations for reform and improvement of democratic quality. As we explain below, we do not assert that these five conceptual dimensions freedom, rule of law, vertical accountability, responsiveness, and equality fully constitute the quality of democracy. The framework we outline below in fact has eight dimensions including as well participation, competition, and horizontal accountability. Others could be identified as well, for example transparency and the effectiveness of representation. The different aspects of democratic quality overlap, however, and we choose to treat these latter two as elements of our principal dimensions. In fact, the more we enumerate distinct dimensions, the denser and more complex the overlaps become. We attempt here to identify some of the ways in which the different elements of democracy not only overlap but also depend upon one another. In identifying these different linkages, we suggest that democratic quality can be thought of as a system, in which improvement in one dimension can have diffuse benefits for others (and vice versa). At the same time however, there are sometimes trade-offs between the different dimensions of democratic quality, and it is impossible to achieve each of them to the maximum degree. This is one respect in which each 2

5 democracy must make an inherently value-laden choice about what kind of democracy it wishes to be. We begin by offering brief definitions of the terms democracy and quality, and we then suggest how these two concepts can be integrated into a multidimensional conception of democratic quality. In the main body of the paper, we define and analyze eight different dimensions of democratic quality. We then reflect further the linkages and interactions among them. Finally, we review how the six case studies employ and illuminate our analytic framework. Democracy and Quality To analyze the quality of democracy, and to seek a good democracy, we must first know what democracy is. At a minimum, democracy requires: 1) universal, adult suffrage; 2) recurring, free, competitive and fair elections; 3) more than one serious political party; and 4) alternative sources of information. 5 If elections are to be truly meaningful, free and fair, there must be some degree of civil and political freedom beyond the electoral arena, permitting citizens to articulate and organize around their political beliefs and interests. 6 In addition, formal democratic institutions should be sovereign in fact, that is they should not be constrained by elites or external powers that are not directly or indirectly accountable to the people. 7 Once a regime meets these basic conditions, further empirical analysis can assess how well it achieves the three main goals of an ideal democracy political and civil freedom, popular sovereignty (control over public policies and the officials who make them), and political equality (in these rights and powers) as well as broader standards of good governance (such as transparency, legality, and responsible rule). 8 Thus, the analysis of a good democracy should exclude hybrid or electoral authoritarian regimes, which, in failing to conduct free and fair elections, fall short of an essential requirement for democracy. 9 By definition, we will find the quality of democracy quite low in defective democracies, which are exclusive in offering only limited guarantees for political rights, or 3

6 dominated in allowing powerful groups to condition and limit the autonomy of elected leaders, and/or illiberal in the inadequacy of their protections for civil rights and the rule of law. 10 We can also expect the quality of democracy to be quite deficient in delegative democracies, which have electoral competitiveness and relative civil and political freedom, but whose officials, once elected, are only minimally responsive to citizen preferences, constrained by other agencies of government, and respectful of the rule of law. 11 A second step in evaluating democratic quality requires a clear definition of quality. A survey of the use of the term in the industrial and marketing sectors suggests three different meanings of quality (each with different implications for empirical research): 1. procedure: a quality product is the result of an exact, controlled process carried out according to precise, recurring methods and timing; 2. content: quality inheres in the structural characteristics of a product, such as its design, materials, or functioning; 3. result: the quality of a product or service is indirectly indicated by the degree of customer satisfaction with it, regardless of how it is produced or its actual content. What is a Quality Democracy? Starting from the definitions above, we consider a quality or good democracy to be one that provides its citizens a high degree of freedom, political equality, and popular control over public policies and policy makers through the legitimate and lawful functioning of stable institutions. A good democracy is thus first a broadly legitimated regime that satisfies citizen expectations of governance (quality in terms of result). Second, a good democracy is one in which its citizens, associations, and communities enjoy extensive liberty and political equality (quality in terms of content). Third, in a good democracy the citizens themselves have the sovereign power to evaluate whether the government provides liberty and equality according to the rule of law. Citizens and 4

7 their organizations and parties participate and compete to hold elected officials accountable for their policies and actions. They monitor the efficiency and fairness of the application of the laws, the efficacy of government decisions, and the political responsibility and responsiveness of elected officials. Governmental institutions also hold one another accountable before the law and the constitution (quality in terms of procedure). With the above in mind, we identify eight dimensions on which democracies vary in quality. The first five are procedural dimensions: the rule of law, participation, competition, and accountability, both vertical and horizontal. Though also quite relevant to the content, these dimensions mainly concern the rules and practices. 12 The next two dimensions of variation are substantive in nature: respect for civil and political freedoms, and the progressive implementation of greater political (and underlying it, social and economic) equality. Our last dimension, responsiveness, links the procedural dimensions to the substantive ones by measuring the extent to which public policies (including laws, institutions, and expenditures) correspond to citizen demands and preferences, as aggregated through the political process. These eight dimensions are elaborated below. 13 Each may vary in the specific form of its institutional expression, and in its degree of development. Capturing and explaining this variation requires indicators that reveal how and to what degree each dimension is present in different countries (and in different models of the good democracy). The resulting empirical data will also make it possible to track trends in the quality of democracy in individual countries over time, including the effectiveness of institutional reforms. 14 The multidimensional nature of our framework, and of the growing number of democracy assessments being conducted in individual countries, implies a pluralist notion of democratic quality. As we note below, there are not only dense linkages but also trade-offs and tensions among the various dimensions of democratic quality, and democracies will differ in the normative weights they place on the various dimensions of democratic quality (for example, freedom vs. responsiveness). There is no objective way of identifying a single measurement framework of democratic quality, right and true for all societies. 5

8 Our analytical framework proposed here partially differs somewhat from other studies on the quality of democracy, such as those of Altman and Perez-Linan (2001) and Lijphart (1999). Both of them develop a quantitative comparative strategy. We emphasize the virtuous combination of qualitative and quantitative measures in the empirical analysis of the phenomenon. Some difference also emerges in our definition of a good democracy and consequently in the dimensions of variation and related indicators. Altman and Perez-Linan refer to three aspects that draw on Dahl s concept of polyarchy (civil rights, participation, and competition). Lijphart includes indicators such as female representation, electoral participation, satisfaction with democracy, and corruption, and these indicators can be associated with our different dimensions (respectively, equality, participation, responsiveness, and rule of law). The institutions and mechanisms of representative democracy are the main objects of the analysis of the quality of a democracy. This is not to ignore direct democracy as perhaps the purest expression of democratic quality, but to acknowledge the actual experience of representative democracies and their real potential for improvement. If the analysis has to be focused on representative democracies, then (vertical) accountability becomes a truly central dimension in so much as it grants individual citizens and organized actors in civil society means of control over politicians and political institutions. In a good democracy, this feature attenuates the difficulties that emerge out of the shift from direct to representative democracy. Vertical accountability is implicitly based on two assumptions from the liberal tradition that highlight the interconnectedness of all of the dimensions explained above. The first assumption is that if citizens are genuinely given the opportunity to evaluate the government s performance, they are in fact capable of doing so, possessing above all a relatively accurate perception of their own needs and preferences. The second assumption is that citizens, either alone or as part of a group, are the only possible judges of their own interests and needs; no third party can determine them. Generally, we accept these assumptions, though they are debatable, and as Dietrich Rueschemeyer 6

9 notes in his essay on equality, those actors with more wealth and cultural capital have disproportionate power to shape citizen policy preferences. Freedom and equality, however they are understood, are necessarily linked to accountability and responsiveness. Indeed, provision of freedom and equality is one standard by which citizens may assess the performance of their political leaders and representative institutions. In addition, an effective rule of law is also indispensable for a good democracy, in part because it guarantees, through the mechanism of a politically neutral and independent judiciary, the defense of citizens rights and of the equal status, before the law and in the political arena, of all citizens. As the next section will explain, freedom, equality, and even accountability are unobtainable ifrespect for law is ineffective or if the elected government cannot govern authoritatively. These are the fundamental presuppositions necessary for identifying and implementing reforms to improve the quality of democracy. Procedural Dimensions of Democratic Quality We are now ready to explore more concretely our eight dimensions of democratic quality in three respects: the empirical definition, the conditions for the dimension to develop and thrive, and the means by which it is commonly subverted. We begin in this section with the five procedural dimensions. The Rule of Law. As Guillermo O Donnell explains, under a rule of law all citizens are equal before the law, which is fairly and consistently applied to all by an independent judiciary, and the laws themselves are clear, publicly known, universal, stable, and nonretroactive. These characteristics are fundamental for any civil order and a basic requirement for democratic consolidation, along with other such cognate features of a constitutional order as civilian control 7

10 over the military and the intelligence services and an elaborated network of other agencies of horizontal accountability that complement the judiciary. More comprehensively, a good (or in essence, a liberal) democracy has a strong, vigorous, diffuse and self-sustaining rule of law in the following respects: The law is equally enforced toward everyone, including all state officials; no one is above the law. The legal state is supreme throughout the country, leaving no areas dominated by organized crime, local oligarchs, or political bosses who are above the law. Corruption is minimized, detected, and punished, in the political, administrative, and judicial branches of the state. The state bureaucracy at all levels competently, efficiently, and universally applies the laws and assumes responsibility in the event of an error. The police force is professional, efficient, and respectful of individuals legally guaranteed rights and freedoms, including rights of due process. Citizens have equal and unhindered access to the justice system to defend their rights and to contest lawsuits between private citizens or between private citizens and public institutions. Criminal cases and civil and administrative lawsuits are heard and resolved expeditiously. The judiciary at all levels is neutral and independent from any political influence. Rulings of the courts are respected and enforced by other agencies of the state. The constitution is supreme, and is interpreted and defended by a Constitutional Court. What further distinguishes and completes a democratic rule of law, O Donnell argues, is that the legal system defends the political rights and procedures of democracy, upholds everyone s civil rights, and reinforces the authority of other agencies of horizontal accountability that ensure the lawfulness and propriety of official actions. To some students of democracy, it may seem odd to begin elucidating democratic quality with the rule of law, as the presence of several dozen illiberal democracies in the world attests to the 8

11 possibility for competitive elections and popular participation to co-exist with considerable lawlessness and abuse of power. 15 Nevertheless, the rule of law is the base upon which every other dimension of democratic quality rests. When the rule of law is weak, participation of the poor and marginalized is suppressed; individual freedoms are tenuous and fleeting; civic groups may be unable to organize and advocate; the resourceful and well connected have vastly more access to justice and power; corruption and abuse of power run rampant, as agencies of horizontal accountability are unable to function properly; political competition is distorted and unfair; voters have a hard time holding rulers to account; and thus, linkages vital to securing democratic responsiveness are disrupted and severed. The above aspects concern in large measure the impartial and efficient application of the law and the fair resolution of disputes within the legal system. Various indicators can represent each and the relevant data can be analyzed on a case-by-case basis using both qualitative and quantitative techniques. Of course, a detailed assessment of judicial fairness and efficiency would be extremely expensive and practically impossible to apply to a high number of cases. However, one can examine across many cases the degree to which public officials are responsible before the law, or instead enjoy relative impunity; the extent of official corruption, as measured (quite roughly and imprecisely to be sure) by indices of perceptions of corruption and of market distortions; the extent of civil liberties violations; the access of citizens to the court system; and the average duration of legal proceedings. The ways in which the rule of law is subverted in a democracy are myriad. For example, politicians may use the law as a political weapon against their political and civic adversaries. 16 Democratically elected leaders may attempt to pack the judiciary (particularly the constitutional court) with political loyalists. There is a growing tendency among individual citizens or economic groups to resort to the courts to assert their own interests (as opposed to strictly defending against illegal actions). The political culture also plays an important role in sustaining or undermining the rule of law. A democratic rule of law is diminished in many countries by the diffuse cultural 9

12 attitude that views the law merely as an impediment to realizing one s own interests, a nuisance to be circumvented in any way possible. In many countries, this attitude extends widely from the popular to the entrepreneurial classes and is captured by the Italian saying, fatta la legge, trovato l inganno (in essence, fraud goes hand in hand with law). The most important conditions aiding the development of the rule of law are the diffusion of liberal and democratic values among both the people and, especially, the elite; strong bureaucratic traditions of competence and impartiality; and the institutional and economic means for fully implementing a rule of law. However, these conditions are not all that common and they are very difficult to create from scratch hence, the weakness of the rule of law in many recently established democracies (and even a number of older ones as well). The best approach is probably to proceed incrementally to build up the independence, capacity, and authority of the entire judicial system. But the literature on rule of law development is sobering:no amount of training and financial resources (including generous external assistance) will suffice unless democratic leaders exhibit both political will and self-restraint. This in turn requires a mobilized and aware civil society, and efficient democratic instruments of competition so that voters can remove public officials who obstruct rule-of-law reforms. Participation. No regime can be a democracy unless it grants all of its adult citizens formal rights of political participation, including the franchise. But a good democracy must ensure that all citizens are in fact able to make use of these formal rights to influence the decision-making process: to vote, to organize, to assemble, to protest, and to lobby for their interests. With regard to participation, democratic quality is high when we in fact observe extensive citizen participation not only through voting but in the life of political parties and civil society organizations, in the discussion of public policy issues, in communicating with and demanding accountability from elected representatives, in monitoring the conduct of public office-holders, and in direct engagement with public issues at the local community level. 10

13 Participation in these respects is intimately related to political equality, because, even if formal rights of participation are upheld for all, inequalities in political resources can make it much more difficult for lower-status individuals to exercise their democratic rights of participation. Thus a fundamental condition for widespread participation in a good democracy is broad diffusion of basic education and literacy, and with it political knowledge of the system of government, its procedures, rules, issues, parties, and leaders. Important again, as a supporting condition, is the political culture, which should value participation and the equal worth and dignity of all citizens. The latter implies as well tolerance of political and social differences, and thus acceptance on the part of all individuals and organized groups of the right of others (including their adversaries) to participate equally, so long as they obey the law. Finally, as we have suggested, extensive participation also requires a rule of law that will defend the right and ability of weaker social groups to participate fully. Participation is subverted and constrained in a variety of subtle and overt ways in democracies around the world. A common subversion comes through the apathy of a citizenry that doubts the efficacy of democratic mechanisms or has become alienated from the democratic process as a result of the low quality of democracy in other respects (for example, corruption, abuse of power, and a lack of competitiveness). Of course, such systemic flaws can, under the right conditions of popular outrage and effective civic mobilization, generate increases in participation, but in the absence of viable reform alternatives, they induce civic withdrawal. Participation may be constrained by conditions of lawlessness and violence that make it risky for citizens to organize, assemble, or even to vote. To preserve their own entrenched privileges, powerful groups may intimidate and victimize the poor, the landless, ethnic, regional, and religious minoritiesin order to suppress their exercise of political influence.participation is most commonly measured by voter turnout rates, but this captures only one aspect of participation in a democracy. No less important are the extent of membership and active participation in political parties, social movements, and 11

14 non-governmental organizations; the frequency of communication with elected representatives and other office-holders; and the extent to which citizens express themselves on public issues. Competition. In order to be a democracy at all, a political system must have regular, free and fair electoral competition between different political parties. But democracies vary in their degree of competitiveness in the openness of access to the electoral arena by new political forces, in the ease with which incumbents can be defeated, and in the equality of access to the mass media and campaign funding on the part of competing political parties. Depending on the type of electoral system, democracies may allow for more or less decisive electoral alternation as well. Here we confront a trade-off within the overall goal of competition. Electoral systems based on proportional representation score well on one element of competitiveness ease of access to the electoral arena and parliament on the part of multiple political parties but at the expense of another element of competitiveness, the ease of alternation of power (or the efficiency of the electoral process), since the presence of multiple parties with relatively defined shares of the vote tends to produce a succession of coalition governments that have some considerable continuity in party composition over time. 17 There is no objective, a priori way to determine from this respect alone which system produces a higher quality democracy, though Lijphart argues that PR does a better job of fulfilling other dimensions of democratic quality, such as the more equal representation of women and minorities. One condition for vigorous competition is the legal and constitutional order. In contemporary democracies, political party and campaign finance is such an important foundation of electoral viability that it is difficult for challenging parties and candidates to compete effectively without some fair minimum in this regard. While there is considerable skepticism about the efficacy of laws that limit campaign spending in part because of the ease with which they are evaded in new and old democracies alike some floor of public funding for significant parties and robust requirements for the full and rapid reporting of all contributions to parties and campaigns do 12

15 seem to promote greater electoral fairness and competitiveness. 18 In first-past-the-post systems, the means by which electoral districts are drawn also heavily shape competitiveness. Where partisan bodies are able to draw electoral districts to their own advantage (as in the United States, where this gerrymandering is performed every decade by state legislatures, with the aid of increasingly sophisticated computer programming), they are likely to do so in ways that will promote partisan and incumbency advantage. As a result, elections to the U.S. House of Representatives have seen the rate of defeat of incumbents steadily decline, and less than ten percent of House seats are competitive. Of course, electoral competitiveness also depends on fairness in access to the mass media, pluralism in media ownership (and viewpoints), some dispersion of economic resources in society, and the enforcement of political rights by an independent judiciary. There is also an important linkage with horizontal accountability, because the most important institutional guarantee of freedom and fairness (and hence competitiveness) in elections is an independent and authoritative electoral commission. 19 Competition can be constrained or subverted through partisan control of electoral administration, not just the drawing of electoral district boundaries but the design and enforcement of rules regarding the conduct and financing of campaigns and the administration of the actual vote. Competitiveness can also be diminished by the actions of parties and other non-state forces that suppress the freedom of different groups and parties to contest for office, or that commit outright electoral fraud. To the extent these subversions become so extensive that declared electoral outcomes do not represent the will of the people, or it proves all but impossible to replace the ruling party through elections, the system falls beneath the threshold of democracy into some form of electoral authoritarian rule. But some regimes straddle ambiguously the boundary between lowquality democracy and semi-democracy. Competition is one dimension of democratic quality that is relatively amenable to quantitative measurement (though no quantitative measure can fully capture the extent of it). Altman and Perez-Linan propose an indicator of the balanced presence of opposition in 13

16 parliament, which has a negative value when the governing party dominates the legislature in terms of seats or when the opposition is so strong that it poses problems for the decisional efficacy of the government. 20 Vertical Accountability. Accountability is the obligation of elected political leaders to answer for their political decisions when asked by citizen-electors or other constitutional bodies. Schedler suggests that accountability has three main features: information, justification, and punishment/compensation. 21 The first element, information on the political act or series of acts by a politician or political organ (the government, parliament, and so on), is indispensable for attributing responsibility. Justification refers to the reasons furnished by the governing leaders for their actions and decisions. The third, punishment/compensation, is the consequence that is imposed by the elector or some other authoritative body following an evaluation of the information, justifications, and other aspects and interests behind the political act. Accountability can be either vertical or horizontal. Vertical accountability is that which citizens as electors can demand from their officials in the course of campaigns and elections, and which political and civil society actors can exercise at moments of political controversy. As Philippe Schmitter explains, in modern democracies, elected representatives play a crucial mediating role in the accountability relations between citizens and rulers. This is especially so in parliamentary systems, where elected representatives can bring down the government, but in all democracies representatives not only receive but help to share and structure citizen preferences and expectations. The electoral form of vertical accountability has a periodic nature and is dependent on the various local and national election dates. Ideally, the performance of the incumbent is reviewed and evaluated, policy alternatives are debated, and the vote either rewards the incumbent party and office-holders or punishes them by voting for an opposition party (or abstaining from the vote). As Schmitter notes, the outcome of accountability may just as likely be reward (reelection) as punishment (defeat), for the most accountable leaders are really those whose conduct and decisions 14

17 conform to citizen expectations (and the law). The dynamics of vertical accountability extend beyond elections and the interplay between voters and their elected representatives, however, encompassing as well the efforts of civic associations, NGOs, social movements, think tanks, and the mass media to hold government accountable in between elections. Smulovitz and Peruzotti refer to this as societal accountability. 22 Political competition and participation are crucial conditions for vertical accountability. If voters are to be able to hold their public officials and ruling parties accountable periodically through elections, they must be engaged, knowledgeable about the performance of their office-holders and the issues, and they must turn out to vote in large numbers. At the same time, vertical accountability requires genuinely competitive elections, in which institutionally strong parties are able to offer programmatic alternatives to the voters, and in which incumbents do not enjoy such entrenched advantages that voters are not able to punish them for poor performance or unwanted policies. In short, political competition and the distribution of power must be fair and robust enough to allow for genuine electoral alternatives at the various levels of government, and to produce some electoral alternation over time, so that incumbents face a credible threat of electoral punishment. The ongoing vertical process of monitoring, questioning, and demanding justification, through the work of non-governmental organizations, think tanks, the mass media, professional associations, interest groups, opposition parties, and other actors in civil society, in turn requires freedom for these groups to function and a rule of law that protects them from intimidation and retribution. In both these electoral and non-electoral respects, vertical accountability therefore requires a strong system of horizontal accountability to sustain the rule of law and thus the integrity of these vertical processes. At the same time, it also depends upon strong and well-established intermediary structures, a responsible, vigilant political opposition, independent and pluralistic mass media that are conscious of their civil function, and a well-developed network of active, informed organizations and associations that share democratic values in other words, a vigorous and vigilant civil society. 15

18 Vertical accountability is subverted when office holderss manipulate the rules and the powers of office to heighten the electoral advantages of incumbency and to handicap the political opposition, or when they extend control over the media and other actors in civil society to the point that the latter fear to question their conduct in office. In the modern era, where the mass media in general and the television media in particular are so vital to the flow of political influence and debate, efforts to subdue or eliminate independent (critical) control of the television media in particular are a common means by which incumbents try to vitiate both the electoral and the civic instruments of vertical accountability. Beyond these rather blatant and deliberate subversions lie more subtle, intrinsic, and pervasive difficulties in securing political accountability from below. Given the opacity and complexity of government, politicians have ample opportunity to absolve themselves of any concrete responsibility. Thus, accountability frequently becomes a catchphrase more connected to the image of a politician than to decisions made or results achieved. Negative outcomes are easily justified by reference to unforeseen events, or by taking advantage of a favorable press or effective image-makers to influence public opinion. At the same time, good results obtained by wise and prudent leadership, even if sometimes at the cost of temporary sacrifices by the governed, might be punished in the next election in reaction against the temporary pain. The very action, often ideological and instrumental, of parties or other opposition actors, or even of powerful media actors, further highlights the difficulty of implementing accountability. The lack of clear distinctions between incumbent leaders and party leaders means that parties, be they of the opposition or of the majority, are hindered in carrying out their role as watchdogs for their constituents. At the parliamentary level, party discipline is considered more important than accountability to the voters and, in most democracies, the parliamentary majority (where it exists) supports the government without really controlling it. Furthermore, there should also be a clear distinction between the responsible leader, either of the government or of the opposition, and the intermediate layers of party actors that range from activists to sympathizers. These latter should 16

19 trigger a bottom-up process that gives direction for how parties should control the government or organize their opposition. Yet recent research on party organization in a few advanced democracies shows a trend toward strong, oligarchic leaders who collude -rather than compete--with other parties. 23 For European citizens, the supranational architecture of the European Union generates additional difficulties in holding national rulers accountable, enabling the latter to shift blame for unpopular decisions onto the EU even if they concern clear-cut issues such as streamlining national administrations or reorganizing state finances to address large national deficits. Governments and politicians justify unpopular actions by claiming that their hands were forced by opposing coalitions in the EU Council of Ministers or in the European Council of prime ministers and chiefs of state, or by votes in the European Parliament. A similar tactic is often used and not without justification by elected governments in developing countries in relation to economic policy constraints imposed by international financial institutions and the World Trade Organization. Horizontal Accountability. Democratic quality also requires that office holders answer for their conduct to and have it reviewed by other institutional actors that have the expertise and legal authority to control and sanction their behavior. In contrast to vertical accountability, the actors are, more or less, political equals. Horizontal accountability is usually manifest in the monitoring, investigation, enforcement and independent functioning of a number of different government institutions (which sometimes overlap in their authority): the opposition in parliament, specific investigative committees of parliament, the various tiers of the court system, including, crucially, the constitutional court, audit agencies, a counter-corruption commission, the central bank, an independent electoral administration, the ombudsman, and other bodies that scrutinize and limit the power of those who govern. 24 The counter-corruption commission is a particularly crucial agency of horizontal accountability in contemporary democracies. To be effective, this body must be charged not only 17

20 with receiving but also monitoring and verifying the assets declarations of the president or prime minister, national-level ministers, members of parliament, state or provincial governors, high-level bureaucrats, major military and police officers, and other elected and appointed public officials above a certain level. (A truly comprehensive effort in a large country will also provide for the commission to have branch offices at the provincial level to monitor provincial and local government officials and legislators). The commission must then have the staff to investigate annually on a random basis some significant percentage of these assets declarations, and systematically, the declarations of the country's highest officials. Scrutiny must be comprehensive if it is to be effective, and if the threat of detection is to be credible. This requires a lot of resources: accountants, investigators, and lawyers trained in the ways that wealth is moved, accumulated, and hidden, along with computer specialists and other support staff to back them up. Not only does a counter-corruption commission need a lot of well-trained staff, it needs to pay them enough to deter temptation and establish a high esprit de corps. Scrutiny, however, is not enough. If credible evidence of wrongdoing emerges, there must be the institutional means to try the suspected offender and impose punishment on the guilty. The single most common and crippling flaw in systems of corruption control is an inability to enforce this function free from interference by the highest levels of government. The counter-corruption commission should have the ability to prosecute officials who have allegedly violated ethics laws independently. One of the most important changes introduced by Thailand's democratic constitution of 1997 was to grant the National Counter Corruption Commission independent prosecutorial authority, even if it means overruling the attorney general. 25 The Ombudsman's office (which may go under many names, including, in South Africa, the office of public protector) receives and investigates public complaints of abuse of office. Members of the public or the press should have a right to indeed be encouraged to bring evidence to the counter-corruption commission if they believe a public official has misrepresented his or her assets or abused their office. But democracies need a supplementary channel of public access to remedial 18

21 government authority if power is being abused and the counter-corruption commission does not pursue its mission, or judges some abuse of power to lie outside its scope of authority. Even if the system is working well, the ombudsman may occasionally come upon evidence that the countercorruption commission does not have, or that reinforces investigations the commission has already opened. And, agencies of horizontal accountability work best when they may be held accountable by one another for failure to do their job. The powers and functions of the office of the Ombudsman vary widely across countries. In some countries, it is simply a mechanism to receive and investigate citizen complaints. The Philippine Constitution, however, gives the Ombudsman an explicit mandate to fight corruption through public assistance, prevention, investigation, and prosecution of suspect public officials, and administrative direction and adjudication. It can order any government official or employee remiss in his duties to do his job, or stop, prevent and remedy improper and abusive acts. In the Philippines, this office need not wait for citizens to come forward, and it can also act on anonymous complaints. Moreover, it may not only refer charges (against lower-ranking officials) to the regular courts, it may also prosecute directly. 26 Thus the office combines the more limited functions of an ombudsman with many of those of a counter-corruption and even supreme audit agency. Independent, systematic audits of public accounts form another crucial link in the web of accountability. Clearly, it is not enough simply to monitor the personal accounts of public officials. A dense, overlapping system of accountability requires that all major government bureaus, agencies, and ministries have their accounts regularly audited, and that they be open to inspection and evaluation of their performance more generally. To conduct these checks, each major government agency or bureau should have its own auditing office and inspector-general. But periodic external audits are also essential. The government should have an office of the auditorgeneral with the authority to conduct external audits on a periodic or random basis, and audit any agency at any time when there is evidence of wrongdoing. One model is the General Accounting 19

22 Office (GAO) of the United States, which is the investigative arm of, and is responsible to, the U.S. Congress (thus giving it substantial autonomy from executive-branch agencies). Horizontal accountability is most effective when it is comprehensive, that is, when the agencies that comprise it interlock and partially overlap in a systemic fashion. Overlapping authority ensures that if one institutional actor fails to perform its duty to expose, question, and punish (and thus ultimately, deter) corrupt behavior, another institution may initiate the accountability process. Interlocking authority means that the different institutions relate to one another in a way that is complementary and reinforcing, so that, for example, an audit agency can uncover fraud, a counter-corruption commission can impose civil penalties for it, and the judicial process can function on its own to press for criminal penalties, while an ombudsman may stand by to investigate and report if other institutions do not work, or to assist and stimulate them in working. As Guillermo O Donnell (who is in some ways the intellectual founder of the term and the field) argues, Effective horizontal accountability is not the product of isolated agencies but of networks of agencies that include at their top because that is where a constitutional legal system closes by means of ultimate decisions courts (including the highest ones) committed to such accountability. 27 The vitality of horizontal accountability hinges most of all on a legal system that, as mentioned above, provides for the exertion of checks and balances by other public entities that are independent of the government, and not competing as an alternative to it. Such a system begins with the law itself (including the constitution), which must provide for strong authority for various governmental institutions to check and monitor one another, while protecting their independence through tenure in office and means of appointment that rise above patronage or partisanship. If the agencies of horizontal accountability (beginning with the courts) are to work, they must have institutional capacity, training, and leadership that is at once capable, vigorous, but also responsible. Like the law itself, the agencies of horizontal accountability can be used as a weapon against political opponents, undermining the credibility of the entire institutional network. 20

23 One type of law that is particularly important regards freedom of information. Malfeasance thrives in secrecy and obscurity. The more that government transactions and operations are transparent and visible, open to scrutiny, the more feasible it is to expose, deter, and contain corruption. For this reason, citizens must have the legal right to request and receive information on all functions and decisions of government that are not a matter of national security or that do not infringe on individual rights of privacy. In the fight against corruption, the public availability of information on government finance, procurement, and contracting is particularly important. Ideally, such information should be posted on the Internet. In particular, all government procurement above a certain (modest) level should be done through competitive bidding that is advertised on government websites. Particularly in political systems with a tradition of corruption and abuse of power, where democratic norms are not deeply rooted, agencies of horizontal accountability, like the electoral administration, need constitutional autonomy. There are a number of possible models for this. In Costa Rica, the Supreme Electoral Tribunal is virtually a fourth branch of government, whose members are elected to staggered six-year terms by a two-thirds vote of the Supreme Court. In India, the electoral commission s independence is protected by explicit constitutional mandate, and by a powerful chairman, who is appointed by the non-partisan president. In some other countries, independence is attained through supervision by a judicial body or by being made accountable to the parliament rather than the executive branch. 28 Subversion of the institutions of horizontal accountability comes most frequently in the appointment process. If the government is able to appoint politically pliable individuals to head these agencies, their potential to scrutinize, question, challenge, and punish may be largely neutralized from the beginning. Executive actors also attempt to undermine horizontal accountability by limiting the legal authority of these agencies, or making them completely dependent for funding on the budget requests of the executive, or preventing them from pursuing 21